Abel Soh
MBBS (S’pore), MRCP (UK)
Consultant Endocrinologist, Raffles Diabetes & Endocrine Centre
Visiting Consultant, Singapore General Hospital
Scope
 The importance of screening
for GDM
 Before HAPO
 The HAPO study and the
IADPSG recommendations
 After HAPO
The Importance of Screening for
Gestational Diabetes
Rates of Gestational Diabetes
<25 years
% 9.0
8.0
25-34 years
>/=35 years
Population-based studies: 4.9 – 12.8%
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1989-90 1991-92 1993-94 1995-96 1997-98 1999-00 2001-02 2003-04
Getahun D, et al. Am J Obstet Gynecol 2008;198:525
Screening of a Disease
The condition should be an
important health problem
Diabetes in Pregnancy is Associated With
Increased Incidence of Adverse Outcomes
Fetal/neonatal
•
•
•
•
•
Maternal
•
•
•
•
•
Spontaneous abortion
Polyhydramnios
Pre-eclampsia
Preterm birth
Cesarean delivery
Congenital malformations
Macrosomia
Neonatal hypoglycemia
Respiratory distress syndrome
Perinatal mortality
Screening of a Disease
The condition should be an
important health problem
There should be an effective
treatment or intervention for
individuals identified through
early detection, with
evidence of early treatment
leading to better outcomes
Results of Randomized Trials for
Treatment of Mild Gestational Diabetes
ACHOIS (2005)
MFMU (2009)
1,000
958
24–34 weeks
24–31 weeks
75-g, 2-hour OGTT
100-g, 3-hour OGTT
FPG <7.8 mM +
2-h PG 7.8–11.0 mM
FBG <5.3 mM + 1-h PG >10.0 mM,
2-h PG >8.6 mM, 3-h PG >7.8 mM
20.4% vs. 3.3%
7.6% vs. 0.4%

No sig. difference
(perinatal death, shoulder dystocia, birth
trauma)
(perinatal death, neonatal hypoglycemia,
 cord C-peptide, birth trauma)
Pre-eclampsia


Weight gain


LGA infants


Shoulder dystocia
No sig. difference

Neonatal fat mass

Not assessed
Number of participants
Time of testing
Diagnosis of mild GDM
Requiring insulin therapy
Primary composite outcome
Crowther CA, et al. N Engl J Med 2005;352:2477-86
Landon MB, et al. N Engl J Med 2009;361:1339-48
Appropriate management of GDM
resulted in reductions in:
• Pre-eclampsia (RR 0.62, 95% CI 0.43–0.89)
• Birthweight >4000 g (RR 0.50, 95% CI 0.35–
0.71)
• Shoulder dystocia (RR 0.42, 95% CI 0.23–0.77)
Hartling L, et al. Ann Intern Med 2013;159:123
Before HAPO
Screening for Gestational Diabetes
Distinguishes between
those who have GDM
and those who probably
do not
Screening of a disease is
usually performed as a 2step process
• Step 1: identify individuals at
increased risk
• Step 2: diagnostic testing
Very important that
screening does no harm
(physical and
psychological)
Screening for Gestational Diabetes
True positive test
False positive test
• Allows for early therapeutic
interventions with potentially
improved outcomes
• Unnecessary interventions
• Inconvenience and anxiety
experienced by the patient
False negative test
True negative test
• Potential of considerable harm
to mother and/or baby
• Potential risk of poor longterm outcomes
• Confirms normal glucose
regulation
• Unnecessary interventions
avoided
Screening for Gestational Diabetes
United States
• 2-step approach:
• Screening glucose
challenge test (GCT)
• Followed by
diagnostic 3-hour
OGTT in screenpositive women
Many other countries
• 1-step approach
• Screening by risk
factors
• Diagnostic 2-hour
OGTT in high-risk
women
Differences in Screening of Gestational
Diabetes Between Countries
Screening strategy
• Universal vs. selective
• 1-step vs. 2-step approach
Test(s) used
Diagnostic criteria
United States – Two-Step Approach
 N = 752 unselected pregnant
women (97% tested in 2nd or 3rd
trimesters)
 Medium: venous whole blood
 Glucose assay method:
Somogyi-Nelson
Glucose screening test
(GST)
• Load: 50g of glucose
• Positive screen: 7.2 mmol/L at
1 hour
Glucose tolerance test (GTT)
• Load: 100g of glucose after 3 days of 250g
carbs followed by an overnight fast
• GDM: at least 2 of the following
• Fasting 5.0 mmol/L
• 1-hour 9.2 mmol/L
• 2-hour 8.0 mmol/L
• 3-hour 6.9 mmol/L
O’Sullivan JB and Mahan C. Diabetes 1964;12:278-85
O’Sullivan JB, et al. Am J Obstet Gynecol 1973;116:895-900
50-g, 1-hour Glucose Screening Test for
Gestational Diabetes
Results
GDM
Present Absent
+
GST
-
15
(TP)
4
(FN)
94
(FP)
639
(TN)
Sensitivity 79%
Specificity 87%
PPV 14%
NPV 99%
O’Sullivan JB and Mahan C. Diabetes 1964;12:278-85
O’Sullivan JB, et al. Am J Obstet Gynecol 1973;116:895-900
Evolution of the O’Sullivan and Mahan
Criteria
The objective of O’Sullivan
and Mahan’s original study
was to devise a scheme that
identified women at future
risk for diabetes after
pregnancy
Predictive value for future
diabetes was validated in a
separate group of women
Subsequent studies show
that GDM diagnosed using
O’Sullivan and Mahan’s
criteria is associated with
adverse perinatal events
O’Sullivan JB, et al. Am J Obstet Gynecol 1973;116:901-904
Gabbe SG, et al. Am J Obstet Gynecol 1977;127:465-69
Evolution of the O’Sullivan and Mahan
Criteria
1979
• The NDDG converted
the original glucose
values (based on
whole blood) to
plasma glucose
equivalent values
1982
• Carpenter and Coustan
converted the original
glucose values (based
on Somogyi-Nelson
method) to equivalent
plasma glucose levels
tested using glucosespecific assays
National Diabetes Data Group. Diabetes 1979;28:1039-57
Carpenter MW and Coustan DR. Am J Obstet Gynecol 1982;144:768-73
The 1999 WHO Criteria
1980 Guidelines
1999 Guidelines
• Used the 75-g OGTT, limiting
the requirement to only 2
plasma glucose readings
(fasting and 2-hour)
• Pregnant women who met
WHO criteria for IGT in nonpregnant adults were
classified as having GDM
• FPG was lowered (from 7.8
to 7.0 mmol/L) for diagnosis
of diabetes
• But WHO remained
ambivalent about the term
IFG (FPG 5.6–6.9 mmol/L)
recommended by the ADA
• Thus, WHO opted to apply
the same criteria to pregnant
and non-pregnant state
World Health Organization. WHO/NCD/NCS/99.2. Geneva: WHO, 1999
The 1999 WHO Criteria
For
Against
• Relatively simple to use
• Studies have shown that
criteria predict both
maternal and fetal
abnormalities in index
pregnancy1 as well as
increased risk for type 2
DM after delivery2
• Criteria were
extrapolated to
pregnant women from
diagnostic cutoffs of the
non-pregnant women
• May not be applicable
to pregnant women
1. Schmidt MI, et al. Diabetes Care. 2001;24:1151-55
2. Kim C, et al. Diabetes Care. 2002;25:1862-68
Before HAPO – Diagnostic Criteria for
Gestational Diabetes
O’Sullivan
(1973)
NDDG
(1979)
Carpenter &
Coustan (1982)
WHO
(1999)
Medium
Whole blood
Plasma
Plasma
Plasma
OGTT
100-g, 3-hour
100-g, 3-hour
100-g, 3-hour
75-g, 2-hour
Criteria for GDM
diagnosis
2 elevated
values
2 elevated
values
2 elevated
values
1 elevated
value
Fasting (mmol/L)
 5.0
 5.8
 5.3
> 7.0
1-hour (mmol/L)
 9.2
 10.6
 10.0
-
2-hour (mmol/L)
 8.1
 9.2
 8.6
> 7.8
3-hour (mmol/L)
 6.9
 8.1
 7.8
-
NDDG: National Diabetes Data Group; WHO: World Health Organization
*O’Sullivan rounded mean + 2 SD values to the nearest 5 mg/dL (0.28 mmol/L) value
O’Sullivan JB, et al. Am J Obstet Gynecol 1973;116:895-900
National Diabetes Data Group. Diabetes 1979;28:1039-57
Carpenter MW and Coustan DR. Am J Obstet Gynecol 1982;144:768-73
World Health Organization. WHO/NCD/NCS/99.2. Geneva. 1999
Before HAPO – Screening and Diagnosis
of Gestational Diabetes
Diagnostic
OGTT
Fasting
1-h
2-h
3-h
Abnormal
values
needed 
Organization
Screening
Screening
method (1-h
threshold)
ADA, 2003
All except
low-risk
50-g GCT
(7.8)
3-h, 100-g
5.3
10.0
8.6
7.8
2
2-h, 75-g
5.3
10.0
8.6
-
2
WHO, 1999
-
-
2-h, 75-g
7.0
-
7.8
-
1
CDA, 2003
All
50-g GCT
(7.8)
2-h, 75-g
5.3
10.6
8.9
-
2
EASD, 1991
Not
specified
Not
specified
2-h, 75-g
5.5
-
9.0
-
-
ADIPS, 1998
All
50-g GCT
(7.8)
2-h,75-g
5.5
-
8.0
-
1
JDS, 2002
All
50-g, GCT
(7.8)
RPG (5.3)
2-h, 75-g
5.5
10.0
8.3
-
2
BSD, 2007
All
FPG (4.7)
2-h, 75-g
-
7.0
-
7.8
1
Risks Factors Used in Selective
Screening
ADA
ADIPS
CDA
 Maternal age



Ethnicity



Glycosuria

Prior GDM


Overweight / obese




Family history of
diabetes







ADA: American Diabetes Association
ADIPS: Australasian Diabetes in Pregnancy Society
CDA: Canadian Diabetes Association
WHO
S’pore




Prior large baby
Prior adverse
pregnancy outcome
NICE





NICE: National Institute for Health and Clinical Excellence, UK
WHO: World Health Organization



No Consensus on Screening and
Diagnosis of Gestational Diabetes
ADIPS
No randomized trials to
compare the use of the
different criteria
ADA, ACOG
WHO
None of the criteria were
based on pregnancy
outcomes
Need for a New Diagnostic Criteria?
3rd International
Workshop-Conference on
GDM (1991)
• Internationally agreed
upon, outcome-based
criteria for diagnosis of
GDM using the 75-g, 2-h
OGTT were highly
desirable
4th International
Workshop-Conference on
GDM (1998)
• ‘There remains a
compelling need to
develop diagnostic criteria
based on the specific
relationships between
hyperglycemia and the risk
of adverse outcome’
Metzger BE, et al. Diabetes 1991;40(Suppl. 2):197- 201
Metzger BE, et al. Diabetes Care 1998;21(Suppl. 2):B161 -7
HAPO – Study Rationale
Overt diabetes clearly
increases the risk of adverse
pregnancy outcomes
What level of glucose
intolerance, short of
diabetes, is associated with
increased risk of adverse
perinatal outcomes?
HAPO Study – Design
Observational study
75-g OGTT at 24–32
weeks
• For all participants
25,505 pregnant
women
• 15 centers in 10 countries
Clinicians blinded if:
• FBG  5.8 mmol/L and/or
• 2-hour BG  11.1 mmol/L
The HAPO Study Cooperative Research Group. N Engl J Med 2008;358:1991-2002
HAPO Study – Outcomes
Primary outcomes
Secondary outcomes
• Birth weight >90th
percentile
• Primary cesarean
section delivery
• Neonatal hypoglycemia
• Cord C-peptide >90th
percentile
• Pre-eclampsia
• Preterm delivery
• Shoulder dystocia/birth
injury
• Hyperbilirubinemia
• Intensive neonatal care
The HAPO Study Cooperative Research Group. N Engl J Med 2008;358:1991-2002
HAPO Study – Maternal Glucose Levels
and Primary Outcomes
Continuous graded
relationships
between higher
maternal glucose
levels & increasing
frequency of
primary outcomes
No obvious
thresholds at
which risks
increased
The HAPO Study Cooperative Research Group. N Engl J Med 2008;358:1991-2002
Affiliated organizations
• DPSG of EASD
• JAPD (Japan)
• ADIPS (Australasia)
• West Coast USA DPSG
• DPSI (India)
• Canadian Special Interest
Group for Diabetes and
Pregnancy
Associated groups
• European Association of
Perinatal Medicine
• Society of Maternal Fetal
Medicine (USA)
• ADA Pregnancy Council
• SAREDIA
IADPSG Consensus Panel
Choose outcomes for
defining thresholds
How much risk is too
much?
Importance of specific
glucose measures
Outcomes
Selected for
Defining Glucose
Thresholds
Strong linear
associations with
increasing maternal
glucose levels in HAPO
At the glucose values
selected to define GDM,
the adjusted ORs were
higher than for any
other outcome variable
Plasma Glucose Concentrations at
Specified Odds Ratios
Mean glucose
level in HAPO
Odds ratio
FPG (mmol/L)
1.5
5.0
1.75
5.1
2.0
5.3
4.5
1-hour PG
(mmol/L)
9.3
10.0
10.6
7.4
2-hour PG
(mmol/L)
7.9
8.5
9.0
6.2
HAPO
participants
meeting
threshold
25%
16.1%
8.8%
*Mean of threshold values for  LGA, body fat, cord C-peptide >90th percentile
IADPSG Consensus Panel. Diabetes Care 2010;33:676-82
First Prenatal Visit
First prenatal visit:
FPG or RPG or HbA1c
FPG  7.0 mmol/L &/or
RPG  11.1 mmol/L &/or
HbA1c  6.5%
FPG  5.1 mmol/L
but < 7.0 mmol/L
FPG < 5.1 mmol/L &
RPG < 11.1 mmol/L &
HbA1c < 6.5%
Treat as for
overt diabetes
Diagnose as GDM
75-g 2-h OGTT @
24–28 weeks’
gestation
IADPSG Consensus Panel. Diabetes Care 2010;33:676-82
At 24–28 Weeks’ Gestation
75-g 2-h OGTT
FPG  7.0 mmol/L
FPG  5.1 mmol/L &/or
1-h PG  10.0 mmol/L &/or
2-h PG  8.5 mmol/L
FPG < 5.1 mmol/L &
1-h PG < 10.0 mmol/L &
2-h PG < 8.5 mmol/L
Treat as for
overt diabetes
Treat as for GDM
No diabetes
IADPSG Consensus Panel. Diabetes Care 2010;33:676-82
After HAPO
Concerns Regarding Recommendations
by IADPSG
Increased incidence
of GDM
Ryan EA. Diabetologia 2011;54:480-86
Cundy T. Diabet Med 2012;29:176-80
Proportion of HAPO Cohort with
Glucose Values  Threshold
Glucose Measure
mmol/L
FPG
 Threshold (%)
Additional
Cumulative
5.1
-
8.3
1-h PG
10.0
5.7
14.0
2-h PG
8.5
2.1
16.1
Total incidence of GDM: 17.8%
(including 1.7% who were unblinded)
IADPSG Consensus Panel. Diabetes Care 2010;33:676-82
Population-Based Reports of GDM Frequency
Based on IADPSG Recommendations
Authors
Country
Subjects
Ethnic/racial distribution
GDM
Frequency
O’Sullivan EP, et al
Ireland
5,500
European 92.9%
Non-European 7.1%
12.4%
23.7%
Black MH, et al
USA
9,199
Non-Hispanic white 7.2%
Hispanic 74.4%
Black 10.1%
Asian 7.4%
Agarwal MM, et al
UAE
10,283
Arab 80.1%
South Asian 15.5%
37.7%
Australia
1,275
Caucasian 89.4%
13.0%
Moses RG, et al
IADPSG Consensus Panel. J Matern Fetal Neonatal Med 2012;25:2564-69
METHODS
• Retrospective cohort study
• 855 pregnant females of
Chinese, Malay, and Asian
Indian ethnicity at high risk of
GDM who underwent 75-g
OGTT between July 2008 and
June 2010
RESULTS
• Prevalence of GDM was
reduced from 28.8% to 21.1%
when the 2013 criteria were
used
• 10.2% were reclassified from
GDM to normal
• 2.6% were reclassified from
normal to GDM
Yew TW, et al. Endocr Pract 2014;20:1064-69
Concerns Regarding Recommendations
by IADPSG
Increased incidence
of GDM
Increased healthcare
resources (workload)
Cost-effectiveness
and health economic
impact
No outcome data of
intervention for
women diagnosed
using IADPSG criteria
Ryan EA. Diabetologia 2011;54:480-86
Cundy T. Diabet Med 2012;29:176-80
The Belgian Diabetes in Pregnancy
(BEDIP-N) Study
Study Information
Study Design
Prospective multicenter observational cohort study
Duration
April 2014 – April 2017
Estimated
Recruitment
2,563
Primary
Outcome
Measures
 Difference in GDM prevalence between the 2-step (50-g GCT
followed by 75-g OGTT) and 1-step (75-g OGTT) IADPSG
screening strategy
 Difference in macrosomia rate between GDM and non-GDM
groups according to the IADPSG criteria
 Glucose tolerance status 3 months postpartum in women with
recent GDM
ClinicalTrials.gov Identifier: NCT02036619
Gestational Diabetes Diagnostic
Methods (GD2M)
Study Information
Study Design
Single-site blinded RCT
Duration
December 2014 – December 2018
Estimated
Recruitment
920
Summary
Primary Outcome
Measures
Secondary
Outcome
Measures
Comparing the Carpenter-Coustan and IADPSG criteria for diagnosing
GDM
Large for gestational age (LGA) infant
 Cesarean delivery
 Maternal composite morbidity (pre-eclampsia, 3rd or 4th degree
vaginal lacerations, postpartum hemorrhage)
 Neonatal composite morbidity (hypoglycemia, hyperbilirubinemia
requiring treatment, clinical jaundice, elevated cord C-peptide,
stillbirth, birth trauma)
ClinicalTrials.gov Identifier: NCT02309138
Early Pregnancy Screening
Rationale
• Particular subgroup might
be at a higher risk (e.g.
prior history of insulintreated GDM)
• Earlier screening and
diagnosis may be
warranted
• Early treatment may be of
benefit in this higher risk
subgroup
Challenges
• Insufficient data to
recommend best screening
test
• Only available evidence
• FPG
• RPG + GCT
More Data Needed
Outcome studies
evaluating impact of
treating maternal
hyperglycemia based
on the new criteria
Cost-effectiveness
analyses of health
economic impact of
implementing the new
criteria
Risk stratification for
women diagnosed with
GDM to various levels
of care
Adoption of IADPSG Recommendations
WHO
Japan
India
The
Philippines
Brazil
Austria
Finland
France
Germany
Greece
Italy
Scotland
Australia
New
Zealand
Canada
Screening and Diagnosis of Gestational
Diabetes (updated May 2015)
Organization
Screening
Strategy
Screening
Method (1-h
threshold)
Diagnostic
OGTT
Fasting
1-h
2-h
3-h
Abnormal
values
needed 
ADA, 2015
All except
low-risk
50-g GCT
(7.8)
3-h, 100-g
5.3 or
5.8
10.0 or
10.6
8.6 or
9.2
7.8 or
8.0
2
2-h, 75-g
5.1
10.0
8.5
-
1
ACOG, 2013
Universal
50-g GCT
(7.5 or 7.8)
3-h, 100-g
5.3 or
5.8
10.0 or
10.6
8.6 or
9.2
7.8 or
8.0
2
WHO, 2014
Universal
-
2-h, 75-g
5.1
10.0
8.5
-
1
CDA, 2013
Universal
50-g GCT
(7.8)
2-h, 75-g
5.3
10.6
9.0
-
1
-
2-h, 75-g
5.1
10.0
8.5
-
1
NICE, 2015
Selective
-
2-h, 75-g
5.6
-
7.8
-
1
ADIPS, 2014
Universal
-
2-h, 75-g
5.1
10.0
8.5
-
1
Summary
Poorly controlled GDM is associated with increased risk of adverse maternal
and fetal/neonatal outcomes
Treatment of GDM (even mild GDM) can reduce the incidence of some of the
adverse outcomes
Before the release of the results from HAPO, there was lack of consensus in
the screening and diagnosis of GDM
HAPO showed continuous association between maternal glycemia and
adverse perinatal outcomes
There remains debate regarding whether the IADPSG-recommended
screening strategy and diagnostic criteria should be adopted by all
However, many countries have adopted the IADPSG-recommended
diagnostic criteria (some with slight modification)